Running DeepSeek Locally on your desktop/laptop with Ollama
DeepSeek, a powerful open-source LLM, can be easily run locally on your desktop/laptop using Ollama. I’m using an M1 MacBook Pro with 32GB. Ollama simplifies the process of running large language models, handling dependencies and providing a consistent interface. This guide will walk you through installing DeepSeek via Ollama, making it accessible with just a few commands.
Prerequisites For Mac
- Homebrew (Recommended): While not strictly required for Ollama itself, Homebrew is highly recommended for managing other dependencies you might need down the line. If you don’t have it, install it from the command line:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
1. Installing Ollama
The easiest way to install Ollama is by going to the Ollama website and picking the binary for your OS. For macOS, I prefer installing things with Homebrew:
brew install ollama
2. Run Ollama
After the installation is complete, open a new terminal window and type the command:
ollama serve
Now the Ollama server is up and listening on http://127.0.0.1:11434
3. Downloading the DeepSeek Model
Ollama manages the downloading of models for you. You don’t need to manually download the DeepSeek model files. Ollama supports a variety of models and quantization levels. Head over to https://ollama.com/search and look for deepseek-r1.

On the deepseek-r1 Ollama library page, you can pick the model you want. The higher the number, the more computing power is needed to get the responses in a reasonable timeframe. I am picking the 8b model that is about 5GB in size.
ollama run deepseek-r1:8b
Once the installation is done, the deepseek-r1 prompt is ready to serve you.

4. Install Docker
Install Docker Desktop on your machine by downloading the binaries from the official website.
5. Installing Open WebUI
Open WebUI is an extensible, self-hosted AI interface that adapts to your workflow, all while operating entirely offline.
Open a new terminal and run the following:
docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main

6. Running Open Web UI from Docker
From your Docker Desktop, click on the Open-webui container and run it. Now the Open-WebUI is ready to serve you with the deepseek-r1 model at http://localhost:3000/

7. Open Web UI on Local machine
From the docker desktop you would know which port the Open Web UI is serving on, here is the screen you would see the very first time. It asks you to enter name, email and password. The login is strictly local, no emails or other information are being harvested. The login is required to protect your Ollama and external APIs from unauthorized access. If you want to run a version without authentication, there is ollama-webui-lite that’s designed to work without a backend (direct browser client -> Ollama API).

Once the admin account creation is done, a release notes screen is displayed, get past that and you are ready to chat with deepseek.

8. Using DeepSeek in your Applications
Ollama also provides a REST API for interacting with models programmatically. This allows you to easily integrate DeepSeek into your applications.
- Start Ollama: Make sure Ollama is running (you might have it running already from the previous step, or start it with
ollama serve
). - Send a request: You can use
curl
or any other HTTP client to send requests to the Ollama API. Here’s an example usingcurl
:curl http://localhost:11434/api/generate -d '{ "model": "deepseek-r1:8b", "prompt": "What is AI" }'
This will send a request to Ollama to generate text based on the specified model and prompt. The response will be in JSON format.
Troubleshooting
- Model Not Found: Double-check the model name you are using with
ollama pull
andollama run
. Refer to the Ollama website for the correct model names. - Port Conflicts: If port 11434 is already in use, you can configure Ollama to use a different port. Check the Ollama documentation for details.
- Resource Issues: Running large language models can be resource-intensive. If you encounter performance issues or out-of-memory errors, consider using a smaller quantized model.
Conclusion
Ollama significantly simplifies the process of running DeepSeek and other LLMs locally. With just a few commands, you can have DeepSeek up and running on your MacBook Pro, ready for interactive use or integration into your applications. Ollama’s API makes incorporating LLMs into your projects seamless and efficient. This approach is much simpler than manually managing models and dependencies, making DeepSeek accessible to a wider audience. Remember to consult the Ollama documentation for the most up-to-date information and advanced configuration options.